J 81 3,] upon Werner^ s Miner alogical Method, 



125 i 



Now, if we consider these two substances for a moment, we 

 cannot but be struck with the great ditferenee of their effects. 

 Gum arabic is an innocent atid even nourishing article of food ; 

 but the poison of the viper is injurious, and even destructive to 

 life. That such opposite properties should eKist in substances 

 absolutely the same we cannot bring ourselves to bt<ieve ; and, 

 cannot avoid concluding that a difference in tlieir composirion 

 actually exists, though the imperfect state oi cht mistry did not 

 permit Fontana to discover it, Nov/ I do not perreive any im- 

 propriety in the hesitation of Werner in such cases ; it is even 

 praiseworthy, and necessary to prevent raineralogicai arraoge- 

 inents from running into confusion. 



2. The first part of Haiiy's definition being tlsus insuii^^ 'ent 

 to enable him to arrange minerals into species, \-. obvious that 

 he must be guided almost entirely by the se ond ciiterioni 

 namely, identity of the integrant niuleeules. 'I his .*cccor.!u)gly 

 5s the criterion by which he is really guided, as tar stony 

 minerals are concerned. All those stones that hr>ve ;h' same 

 primitive form of crystals he considers as belonging to sanie 

 species. Now granting that all minerals are fourid in a t iv-' d- 

 lized form, and granting it to be an easy matter to dettnnine 

 the primitive crystalline form with marbtniatical accuracy, siIU 

 I affirm that the shape of the crystal ims been pitched on to 

 determine the species without sufficient consideration. The 

 atoms of which these primitive crystals are composed have 

 doubtless all of them a determinate form; th*y may, therefore, 

 (unless we suppose them all spheres) unite by different faces, and 

 of course form crystals of different shapes, not reducible irom 

 each other by any conceivable law : so that it is at least possible 

 .that the two parts of Haliy's definition may become inconsistent 

 with each other. Identity of chemical composition may exist 

 along with a diversity in the shape of the integrant molecuies. 

 Accordingly two instances of this kind have already occurred. 

 Calcareous spar and arragonite, and ruthile and octaludrite, as 

 far as chemical analysis goes, are absohiTely ideniical ; yet they 

 differ in the form of their integrant molecrdes. Mr. Chenevix 

 acknowledges the existence of the first of these exceptions, but 

 he was not aware of the second. He even plumes himselt upon 

 the importance of a theory against which only one valid excep- 

 tion could be stated, not considering that one valid exception 

 overturns a theory just as effectually as a thousand. 



But Haiiy's rule for forming species is attended with other 

 inconveniences of a very important kind. It freCjU< ntly unites 

 together minerals which posses^ very different and disiinci cha- 

 racters; nor is it impossible that it may separate minerals whose 

 pther characters very nearly coincide. What minerals, for 

 example^ can be better distinguished from each other than 



